Cardiovascular disease (CVD) patients were targeted in this study to assess the performance of the proposed methodology in segregating patients having stenosis and also identifying the principal diseased coronary artery using PPG and ECG signals. Three major arteries left anterior descending coronary artery (LAD), right coronary artery (RCA), and left circumflex coronary artery (LCx) were focused in distinguishing the stenosed coronary artery. In this prospective cohort study, we considered coronary angiogram as the ground truth.
PPG and ECG signals were acquired simultaneously from CVD patients visiting cardiology department of Medical College and Hospital. Specific inclusion and exclusion criteria were followed during data collection as per the Institutional Ethical Committee. After pre-processing the signals, dicrotic notch region of PPG and S-T segment of ECG, within each cardiac cycle was extracted as a template. A new fused segment was generated from two templates by a proposed algorithm. Utilizing statistics on three templates we defined the term Coronary Health Index (CHI) to evaluate the status of coronary arteries. Setting CHI thresholding values, healthy and stenosed arteries were differentiated. Using CHI values from patients with stenosis, the classification of LAD, RCA, and LCx was performed using Graph Attentive Convolution Network (GACN).
Among 408 CVD patients, 256 had occlusion in either LAD or RCA, or LCx. Binary classification among presence and absence of stenosis was carried out with 0.92 accuracy, 0.91 recall, 0.91 precision, 0.90 specificity, and 0.92 F-score. Identification of stenosed artery was measured with Kappa score (0.89) and Youden’s J statistic value (0.84). Ablation studies were made to design the GACN. Comparison with other networks and state-of-the-art was conducted in terms of AUC(0.93) and AP(0.92) values from ROC and PRC curves, respectively.
The CHI derived from the PPG and ECG signals could be able to study stenosis in non-invasive, easy, and cost-effective manner.